Seemed simple but actually turned out to be quite complex.
ND_MODE notifier was being misused, changed to ND_TOOLSETTINGS
Note: Although the fix itself doesn't involve this much, debugging uncovered
that listeners can't rely on notifier->swinid to be correct, since notifiers can
now be added from RNA without a region context.
Fileselect handlers are added to window's modalhandlers, and contain references to
other areas/regions/etc... Still seems a little bit dodgy, but fixes file selectors for now.
Fun bug, took me the better part of the day to track down.
Happens because maximizing swaps spacedata lists between the old area and the newly created maximized area (this one being empty) while ui handlers are still hanging with references to the first area (then trying to access spacedata when handled). And then only if a maximizing operator was run before the UI realign timer event from the previous maximize was handled (fun, I told you).
After discussion with Matt on irc, we decided the best way to deal with that was to remove ui handlers that reference areas of a screen that is no longer used. That solution reflects the fact that the bug is more general that the reproducing steps would lead to believe. There's also absolutely no reason to run UI handlers on invisible areas.
- 2 fingers scroll (MOUSEPAN / GHOST_kTrackpadEventScroll event) pans/scrolls the view
- 2 fingers pinch (MOUSEZOOM / GHOST_kTrackpadEventMagnify event) zooms the view
And in 3D view:
- alt + 2 fingers scroll rotates the view
- 2 fingers rotation (MOUSEROTATE / GHOST_kTrackpadEventRotate) orbits the view.
The implementation uses a new GHOST event type: GHOST_kEventTrackpad, that is then dispatched as Blender MOUSEPAN, MOUSEZOOM
or MOUSEROTATE events.
This is currently fully implemented for OSX (GHOST Cocoa fires the new events), with auto-detection of the source peripheral, so that a regular mouse still sends MOUSEWHEEL events.
introduced one in wmWindow.last_click_time. Moved this to the wmEvent
struct, which now no is in DNA, was needed for RNA wrapping but not
needed anymore.
Takes into account the hierarchical structures of keymaps as well as wildcards (KM_ANY) in event definitions, user remaps (emulate numpad, action/select mouse buttons, ...) and event values that overlap (click, press and release)
For now, doesn't do anything other than print conflicts in the console.
As a result, I cleaned up a lot of keymaps that had double definitions, moved some keymap items in more appropriate places, fixed wrong definitions and removed kmi that were added for testing a long long time ago.
Out of all the remaining conflicts, after removing obvious non-issues, here's what remains: http://www.pasteall.org/9898
1) Double Click takes priority over a second Click (Click will still be send if not handled)
2) The first Click being handled doesn't stop the following Double Click
* Added ANIM_list_elem_update and ANIM_id_update functions to call when
changing animation curves, which will then call the RNA property update
functions for those curves.
This is to replace just calling DAG_id_flush_update, that may not always
be the right thing to do, and doesn't send proper notifiers for redraw.
Still only used/usable when transforming in the graph editor, not sure
how this do this with NLA for example, .. needs to be improved.
* Added function wm_data_handle_update function to contain the object
update function, and also added scene animation update there. Actually
it should be doing all datablocks, this makes it work for sequencer.
Joshua, do you agree this is the right direction to go in? I can revert or
change the code if you think it should be done differently. Mainly wanted
to get this working well for sequencer now.
Now, temp screens are hidden from being accessed directly, with a new 'Back to Previous' button appearing in place of the screen menu when (for example) fullscreen render image areas are present. Window type menus also get disabled here too, to prevent things from getting too mixed up.
modal operators should return RUNNING_MODAL|PASSTHROUGH for unhandled events to be able to receive clicks correctly (this needs to be fixed for other modal operators).
Maybe it's time to have "handled" flag in event instead.
Recent click event changes moved variable up one level too far.
Also fixes use of timer event for fly operator, it should always check for
which timer it is, not really related to this bug.
Leftmouse+Ctrl to extrude now mapped to Click instead of Release. Release was used to avoid conflict with lasso, but it isn't safe with modal operators that use Press to confirm (subsequent Release then extruded). Click is semantically closer to what we want here.
This may result in cursor being grabbed twice (though we don't have any macro that can do this at the moment). If this is a problem, a check can be added.
Make macros work with more than one modal operator and mix of modal and invoke/exec
As a proof, brought back loopcut + edge slide in a single macro operator called Loopcut and Slide, replacing Loopcut as assigned to Ctrl-R.
Fixing/Missing RNA properties:
- "virtual" property for "ANY" modifier for keymapitem
- modal property for keymap
- Look up modal_items in usermaps too
Lazy init usermaps needs to init modal_items too.
New function to initialize a user keymap (fill in modal_item and poll pointers).
Operator modal keymaps now look up if there's a user defined keymap that overwrites it.
Full Event UI buttons now show "Any" when modifier is set to that (instead of listing all of them).
Note: Having the modifiers as boolean still doesn't express the full breath of values possible for them. There is commented code in this commit to represent them as enum, which would solve this, but make the keymap editor more cryptic.
Auto save is now working again in 2.5. It will also remember now what
the location of the original file was when recovering it, so that
library links still work and saving the restored file does not save to
the temp directory. There is also a new Recover Auto Save operator
which will open the filebrowser in the temp directory and show the
auto saved .blends.
Implemenation Notes:
* Timer storage was moved from window to windowmanager, so we can have
windowmanager level timers too now, doesn't make sense to have
autosave timer attached to a particular window.
* FileGlobal now has a filename field storing where the file was saved.
Note that this is only used when loading a file through the recover
operators, regular file read doesn't use it, so copying the quit.blend
manually over the original file will still work as expected.
* Jobs timer no longer uses operator now, this seems more like an
internal thing, changing keymaps should not make it possible to break
the jobs manager.
* Autosave is postponed by 10 seconds when a modal operator is running,
e.g. transform or file browsing.
* Moved setting G.sce in setup_app_data before depsgraph updates, these
can use the filename for pointcaches.
- Use an enum for grab modes rather then boolean options.
-- GHOST_kGrabNormal: continuous grab userpref disabled
-- GHOST_kGrabWrap: wrap the mouse at the screen bounds *
-- GHOST_kGrabHide: hide the mouse while grabbing and restore the mouse where it was initially pressed *
GrabWrap is nice for transform and tools where you want some idea where the cursor is, previously I found both restoring the mouse at its original location and restoring at a clamped location was confusing with operators like transform, wrapping is not ideal but IMHO the best of a bad bunch of options.
GrabHide is for numbuts, where restoring the mouse at the initial location isnt so confusing.
It will check if either the operator or operator type flags are set on top of the user preference before grabbing the pointer.
I've set that flag for 3d view navigation operators, others should be set too (no transform, I'll deal with that one).
Keymaps are now saveable and configurable from the user preferences, note
that editing one item in a keymap means the whole keymap is now defined by
the user and will not be updated by Blender, an option for syncing might be
added later. The outliner interface is still there, but I will probably
remove it.
There's actually 3 levels now:
* Default builtin key configuration.
* Key configuration loaded from .py file, for configs like Blender 2.4x
or other 3D applications.
* Keymaps edited by the user and saved in .B.blend. These can be saved
to .py files as well to make creating distributable configurations
easier.
Also, user preferences sections were reorganized a bit, now there is:
Interface, Editing, Input, Files and System.
Implementation notes:
* wmKeyConfig was added which represents a key configuration containing
keymaps.
* wmKeymapItem was renamed to wmKeyMapItem for consistency with wmKeyMap.
* Modal maps are not wrapped yet.
* User preferences DNA file reading did not support newdataadr() yet,
added this now for reading keymaps.
* Key configuration related settings are now RNA wrapped.
* is_property_set and is_property_hidden python methods were added.
of the main reasons for passing along context, while actually
they don't need much context at all. Might be removed again,
but would like to have this especially for RNA API functions.
- Useful for dragging buttons to the far right when theyd otherwise hit the screen edge.
- Useful for transform though probably NOT what you want when using the transform manipulator (should make an option).
- When enabled, number buttons use this as well as a different conversion of mouse movement
float numbuts: mouse 1px == 1-clickstep
int numbuts: 2px == 1 (tried 1:1 but its too jitter prone)
details...
- access as an option to GHOST_SetCursorGrab(grab, warp)
- Currently all operators that grab use this, could be made an operator flag
- only Ghost/X11 supported currently
(Mostly for very early testers)
Cocoa uses coordinates with y=0 at bottom : updated wm_window.c and wm_event_system.c for COCOA build to avoid double conversions in response to mouse move events and GHOST_getCursorPosition
Known limitations:
No fullscreen support
Font issue in preference panel
libSDL uses some Carbon functions
one active. Now there's a function to get the pointer + property from
the UI, just like for the animation operators.
Also two fixes for fileselect events, regions are now preserved so that
context is restored to the old region, and the cancel callback is called
when the operator is cancelled.
Main Feature:
* It is now possible to choose which AnimData block is the 'active' one for editing, and/or select them too. AnimData blocks are generally the dark blue and lighter-blue expanders (i.e. Scene, Object, Camera, Lamp, Curve, Armature, etc.)
* Objects are no longer selected/deselected when AKEY is used to toggle selection of channels. This was getting a bit annoying.
* Following on from selection of AnimData blocks, it is now possible to select/make active an AnimData block in the animation editors, and change the active action for that block via the 'Animation Data' panel in NLA Editor's properties region.
--> Be aware that user-counts are not totally handled correctly there yet, so some funky behaviour might be seen...
--> It is possible to assign a new action, or to assign an existing one, allowing to switch between actions as in the past with Actions/IPO Editors...
Other tweaks:
* Some code tweaks towards making the 'Euler Filter' feature for Graph Editor working sometime soon
* Added some backend code for snapping the values of keyframes to a single value. Still need to work out some UI for it though.
* Shuffled the code for ACT_OT_new() around, and removed the poll() callback so that it worked in NLA too.
* Fixed some more notifier bugs with deleting bones and a few other editmode operations for Armatures.
With mouse events event->val started as 0/1 for press/release but later the tweak function made LMB and RMB zero value into KM_RELEASE, somehow MMB didnt get used by the tweak function so was left at 0 and the modal keymap function failed when comparing MMB Mouse ups.
now initialize event->val as KM_PRESS/KM_RELEASE
- rename WM_OT_ten_timer to WM_OT_redraw_timer
- added iterations argument to run more then 10 times (10 is default still)
- use report api rather then always calling a popup directly.
- added a new test that draws every region without swapping.
- dont show the info popup when operators are called from python.
- operators called from python now print reports, useful with the interactive console.
eg.
>>> bpy.ops.wm.redraw_timer(type='DRAW_WIN', iterations=300)
Info: 300 x Draw Window: 4168.56 ms, average: 13.8952
text editor creates squares on some systems. Based on info from
Martin, it appears the keymodifier is being set when it shouldn't.
I think this is happening become some systems may be generating
KM_PRESS events without a matching KM_RELEASE?
Also ignore ascii values 1-32 now instead of 14-32, not sure why
they were included now in 2.5 because they were not in 2.4, but
I don't see a reason to do it. This fixes squares when pressing
e.g. ctrl+b or ctrl+n.
